15

The Satpati handaxe site and the Chabeni uniface site tn southern Nepal

A handaxe site in tectonically folded alluvial sandstones at the Frontal Himalayan Thrust zone

by Gudrun Corvinus, Kathmandu/Erlangen

I.lntroduction

In January 1991 a second site with handaxes and bifaces could be recorded in Nepal in interesting and particularly important context. This was during a short fieldtrip together with Japanese geologist collegues, who were doing geological mapping in the area, in order to clarify and check the location of the Himalayan Frontal Thrust as weil as the geology around the thrust in the area west of the Narayani River emergence into the Terai Plain in the Nawal Parasi District in western Central Nepal. The Japanese team had worked in this sector of the Nepal Siwaliks for a number of years and produced a detailed geological map of this area (Tokuoka et al. 1986, 1988). The author's own Siwalik survey (Biostratigraphical investigations in the Siwaliks in Dang Deokhuri District, Corvinus 1988, Corvinus and Nanda 1994) led besides substantial palaeontological findings to the discovery of a great number of prehistoric sites in the foothills of the Himalayas and in the Dun valleys of Dang and Deokhuri, which for the main part are of later Pleistocene and of Holocene age. A first site of handaxe occurrence was discovered in 1990 in basal g ravels in Dang Dun valley (Corvinus 1991, 1993) which is of particular interest, as it is a first handaxe site well wirhin the Himalayan mountains.

Acheulian handaxe-cleaver industries are very common in the Indian peninsula to the south of the Gangetic Plain, and also, but much less frequent, in Northwest lndia in Jammu and Kashmir. But north of the Gangetic Plain no handaxe occurrences had been recorded previously in the Himalayan foothills, nor in the Gangetic Plain itself, as the upper, exposed sediments of the Gangetic basin are much too young to contain remains of early palaeolithic industries. lt was therefore of particular interest to have found palaeoliths from the Gangetic basin which have become uplifted and exposed through tectonic

movements at the Himalayan front. The site described here, lies 7 km NW of Tribenighat where the Narayani River enters the Terai plain , near the village of Satpati in Nawal Parasi District in Lumbini Zone at the foot of the Siwalik Hills at 27°28 ' latitude and 83°52' longitude.

II. Geology and site situation.

The foot of the Siwalik rangerunsherein a NW-SE direction with Lower to Middle Siwalik deposits of the Arung Khola Formation (Tokuoka et al.1988) bordering the alluvial plain and seperated from it by

16

E

~ f j 6

83°52°

Gudrun Corvinus

r-::::::1 t::.iJ r=:=3 l::::3

coarse gravels

}Satpat l Hili depOSitS

s1ltstones sand- and

silts of the Chabeni

Siwaliks

0 Mahalbari

Fig. 1. Geological map of ehe Saepaei area. + indicaees ehe palaeoliehic siee and x ehe Chabeni siee.

SW ~ -----,> NE Salpali Hill (305 m)

~ ~ >' .~

"' l I

low terrace

0.5 1 km 1.5km 2km 2.5 km

Fig. 2. Geological profile of ehe Saepaei Hili .

27° 29°

300

200

100

The Satpati handaxe site and the Chabeni ttniface site in southern N epal 17

the Himalayan Frontal Thrust (HFT), or Frontal Churia Thrust (FCT) as the Japanese call this frontal thrust. Along this thrust the Siwalik deposits override the younger Pleistocene and Holocene alluvium of rhe Gangeeie Plain, which are estimared to be a few thousand meter thick.

At this locality, however, north of Satpati village, the HFT bifurcates into rwo strands, according to Tokuoka and Takayasu (pers. comm.), around an oblong hill of alluvial deposits of sandstones and gravels, which in Takayasu's opinion are of post- Siwalik age, belanging to the Gangetic Plain alluvium. The deposits at this place are included into the tectonic activities of the HFT and are folded and rilred. That the sediments of the oblong hill, called hereafter the Satpati Hili, in facr do belong to the Gangetic alluvium could be proved by the Eindings of handaxes in situ in these deposits, which was really an unexpected discovery. The Satpati Hili rises, for abour 2 km, 200 m above the alluvial plain of Satpati , which lies at an elevation of 106 m above mean sea Ievel (Fig.1). The upper part of the hill consisrs of a colluvial deposit of very ill-sorted, coarse blocks and boulders of Siwalik sandstones. Below it rhe hill exposes soft, light-grey, little consolidated, micaceous beds of sandstone, siltstone and gravel. They form a small anticline (Fig.2). Irs northern flank dips to the NE and its southern flank to rhe SW. The NE dipping colluvial deposit at the top of the hill seems to overly conformably the micaceous Sandstones and gravels (Takayasu, pers .comm.). The deposirs of the Sarpati Hili will be called hereforth the Satpati Beds.

On the northern part of the Satpati Hili the Sarpati Beds are overidden by the Himalayan Frontal Thrust with Siwalik rocks of alternations of sandstones and mudstones, conforming to the Arung Khola Formation (Tokuoka 1988), and the Chor Khola Formation (Corvinus 1988, 1993). On the souehern side of the Satpati Hili the folded and uplifted Satpati Beds are overiding rhe recent alluvium of rhe Terai along a sourhern-most thrust (Fig.1). The Sarpati Beds of rhe oblong Hili of Sarpati , which have been affecred by the tectonic activiries of the HFT are certainly younger than the Siwalik rocks and belong to the posr-Siwalik alluvial sediments of the Gangetic Plain.

In order to find evidences for the assumption that the Satpati Beds are indeed of post-Siwalik age, thus belanging to the Gangeeie alluvium, the author searched these deposits for any signs of fossils or stone tools. With great surprise and satisfacrion we were able to find not only a few fossil bones together wirh a molar of Bos cf. namadicus (Badam, pers.comm.), eroded out from the micaceous sandstones, 30 m up rhe hill, but also a number of early palaeolithic stone tools, particularly a few bifaces of undoubredly early palaeolithic age (Fig.3). The stone tools were found on the hill slope at the same place as rhe Bos tooth, and in a narrow gully on the hill slope. They are definitely deriving from the Satpati Beds, as some of them also show calcrete adhering to them. The toollocality on the hill, called here the Satpati site, is situared exactly at the crest of the small anticline, where a lower gravel layer of mainly weil- rounded quartzite cobbles in association with Siwalik sandstone cobbles is exposed on the hill slope. The palaeoliths are in all probability derived from this g ravel layer and maybe from the sandstone just overlying this gravel, as they are all found wirhin the direct vicinity of this gravel and sandstone. Besides that, a butt part of a fresh biface was found in situ (with 80 % probability of being in situ), in a NE

dipping sandstone in the lower part of the Satpati Hili, exposed on the hill a little further east of the main handaxe locality. It is completely fresh and unwearhered. Further up the hill an upper g ravel layer is exposed above the sandstones. Bur this gravel differs from rhe lower one in thar it contains mainly pebbles and smaller cobbles of Siwalik Sandstones and very few pebbles of quartzite (Fig.2). Ir is unlikely that this deposit is the rool-bearing deposit.

These findings of early palaeolithic tools wirhin folded sandstones and gravels at the Himalayan Frontal Thrust is of particular interest for several reasons:

1). It is the second, rare occurrence of early palaeolithic tools in Nepal, rhe other being the site of Gadari in rhe Dang Dun valley in basal gravels of rhe Dun alluvium in the wesrern Siwaliks (Corvinus 1991).

18 Gudt'lln Corl'imu

Fig. 3. The handaxe sire (x) on ehe Sarpari Hili.

Fig. 4 The arpari handaxe site. The original horizon is ar ehe place where ehe man stands. A handaxe is een below,

ar ~, eroded out in the gully.

The Satpati handaxt Jitt ar1d the Chabeni 1111ijace Jile i11 wuthem Nepal 19

Fig. 6. The main gully of ehe Sacpaci Hili, into which arcefaccs have been washed down.

20 Gudmn Corvinus

2). There have not been any previous incidents along the Indian and Nepali Himalayas of finding definite palaeolithic implements wiehin folded sediments of the Himalayan tectonics. Only recently Rendell and DenneU (1985) reported from Pakistan a few bifaces in situ in Upper Siwalik contexts that can be dated wich some confidence to between 700 000 and 400 000 years b.p.

3). The finding of probably late Middle Pleistocene bifaces in these folded alluvial deposits g ive age indications for the movement of the Himalayan Frontal Thrust along the foot of the ~imalayan

mountains, bordering ehe Gangeeie Plain. The deposits are younger than the Upper Siwaliks, but older than the young, recent Terai terrace deposits which they underly. They seem to belong to older Gangeeie molasse deposits, buried by about 200 m of younger alluvial and colluvial deposits.

The tool-bearing deposits have become exposed only due to the tectonic events and uplift at the HFT. The palaeoliths are in situ in rhese deposits (Fig.4) and indicate a probable late Middle Pleistocene age for these deposits which ehe rooth of Bos cf. namadicus confirms. Palaeolirhic man must have lived here at the foot of the rising Siwaliks along a sandy and gravelly river bank which became buried by at least 200 m of younger alluvial sediments. They became exposed again, after the deposition of the 200 m of younger alluvium, by the last tectonic events along the Himalayan front which could have happened only very recently in late Pleistocene times.

Furthermore terraces have formed along ehe Himalayan front in the Terai at this place below the Satpati Hill and the Siwaliks. The villages of Satpati and Chabeni are situated on the low, recent wide alluvial plain of the Terai of an elevation of ca.100 m. above mean sea Ievel. 20 m above this low terrace plain a higher, older terrace has formed, just north of the village of Chabeni. lt consists of alluvial silts, topped at places by narrow colluvial fans of Siwalik sandstone blocks and boulders. These colluvial blocks are, in my opinion, of much younger age than the thick colluvial sheet of Siwalik sandstone blocks overlying conformably the Satpati sandstones on the top of the Satpati Hill.Such colluvial fans are very common along the foot of ehe Siwalik hills and seem to be sudden, small flush deposits (as in the case of the Chabeni terrace) or !arger Iandslide deposits from the hills into the foreland (as probably in the case of the colluvia on top of the Satpati Hili).

This higher terrace of Chabeni has yielded another very interesting prehistoric occurrence of an occupation site ofyet unknown but certainly Holocene age. The Chabeni site contains a large nurober of rough flakes and waste debris together wich unifacial adzes, unifaces and unifacial scrapers, which show surprising affinities to Hoabinhian unifacial tools from South East Asia. The arrefacts are in the process of being eroded out from the top part of the alluvial silts of the higher Chabeni terrace. The Chabeni terrace was thus formed just prior to the occupation of the Chabeni cool people. The terrace is probably only a remnant of a formerly much more extensive terrace plain and is at present rapidly being eroded away by the recent streams flowing out from the Siwaliks (Fig.S).

Ill. The cultural rema1ns

1 . The Satpati ea rly palaeolithic assemblage

Altogerher 18 arrefacts have been collected from ehe Satpati deposits of ehe Satpati H ill together with a tooth of Bos cf. namadicus as well as unidentified limb bone fragments with a matrix of micaceous sandstone in its cavities. a. Location and situation

The Satpati handaxe site is marked with x in ehe map (Fig.l). All the artefacts come from the slope and the gullies of the central part of the Satpati Hill at a cliff face called Satpati NE. T he first handaxe (Nr. 1) was found in the cobbly bed of ehe small, dry srream or khola * which drains from the Satpati NE cliff,

* Khola means here: a small, dry stream bed which carries water only during rains.

The Satpati handaxe site and the Chabeni uniface site in southern N epal 21

(Fig.6) and cuts the Satpati deposits of micaceous sandstones and gravels. The cobbles in rhe little stream at rhis place are mainly of Siwalik Sandstones but also of less frequent quartzite cobbles, which derive from the lower gravel layer wirhin the sandstones, silts and gravels exposed in the Satpati NE cliff. Ir is uncertain whether rhe handaxe has derived from the sandstone or the gravel. Ir has no matrix adhering and is somewhat rounded , but not broken and it is more likely to have derived from the sandstone.

Another very well-made, bur anciently broken handaxe (Nr. 2) was found in a small wesrern gully which drains from the NE cliff deposits inro the cenrral stream of the Satpati Hill, some 5 m above the bed of the central khola bed (Fig.4), amongst cobbles of quartzite and Siwalik sandstones, which derive from the lower gravel bed. From the same gully, some 10 m higher than handaxe Nr. 1 comes another large biface (Nr. 13), made of white quartzite, wich a particularly well-made rounded poinr, which was found in May 94.

A third, small and very fresh biface (Nr. 3) was found higher up on ehe hill, some 40 m above the cenrral Satpati khola bed. Ir seems to have been freshly eroded out from the Satpati sandstone above the lower gravel layer of ehe Satpati NE cliff. Nearby this biface we found the molar of Bos cf. namadicus, wearhered out from the Satpati sandstone at this place. Ir has a sandstone matrix in its cavities.

Two large, fresh flakes (Nr. 4 and 5) and two bifacial artefacts (Nr. 6 and 7) were found here, too, on the slope jusr below the upper gravellayer, some 40 m above rhe central khola bed. Nr. 5 and 6 both have calcrete crusts adhering to their surfaces and are very fresh and completely unweathered. They seem to have only recently been derived from wirhin the micaceous sandstone beds at this place. Thesesandstone beds have a NE-dip and belang therefore already to the northern flank of ehe small anricline of the Satpati deposirs.

A very fresh, rough, probably unfinished biface (Nr. 10) was found near the orher bifaces on rhe same slope, some 20 m above the cenrral khola bed and seems to have recently been eroded out from wirhin the same deposits.

An endchopper (Nr. 15) was found in May 94 on the slope in direct conract with the lower quartzite gravel outcrop, 30 m above the cenrral khola bed. Ir has calcrete adhering to its surface and seems to have come out from ehe gravel. Three small flakes (Nr. 16-18) were found at the same place tagether with 3 fragmenrs of fossil bones and there seems no doubt that they, too, are derived from the same g ravel.

Two more artefacts, a pick and the butt part of a biface (Nr. 8 and 9) were located in a small stream bed just west of the main locality. They must have been eroded out from ehe same deposits as ehe others, but were washed down the slope to ehe west instead of to the east as the others.

Nr. 11, a slightly rounded bifacial endchopper or unfinished biface, has been recovered from ehe cobbly cenrral khola bed, just upstream of the mouth of the small gully with the bifaces Nr. 2, 3 and 13 . In the same srream bed the blue biface (Nr. 14) was found recently in May 94.

After climbing over the sreep ridge to rhe nexr eastern cliff of the Satpati deposirs the most convincing in situ implemenr, (Nr. 12), was extricated from a NE-dipping sandstone bed wirhin the lower part of the succession of Satpati Beds. A few meters above it, it is overlain by the upper g ravel layer, which is the

conrinuation of the upper gravel layer described previously. I t is the broken butt part of a very fresh, unweathered biface with calcrere adhering to irs surface. Summing-up the occurrences:

Nr. 1,11 and 14 were found in the cobbly, sandy bed of the cenrral khola at the Satpati NE cliff; Nr. 2, 10 and 13 were found on the slope, 5- 20m above the cenrral khola bed, at Satpati NE. Nr. 2 is

5 m, Nr. 13 is 15 m and Nr. 10 is 20 m above rhe cenrral khola bed; Nr. 3, 4, 5, 6 and 7 are from the slope at the Satpati N E cliff, just below the outcrop of the upper

gravel layer and above the lower gravel layer , about 4o m above the cenrral khola bed; N r. 15, 16, 17 and 18 are found at the outcrop of the lower gravel bed, about 30 m above ehe cenrral

khola bed;

22 Gudmn Corvintts

Nr. 8 and 9 are found in the adjoining khola bed, west of the main artefact location, washed down from ir towards the west instead towards the east like the orhers;

Nr. 12 is found in situ in the sandstone between the lower and the upper gravellayer, but a little to rhe easr of the main locality. b. Description of arrefacrs Bifaces from Satpati:

There are 7 bifaces, of which Nr. 2, though broken, is rhe finest specimen. lt is rhe lower part of a !arge handaxe with only little cortex left near the butt (Fig. 7). It has a flat biconvex section and is rrimmed bifacially by shallow, primary flakes and deep and shallow srepflakes and has a fine retouch along both lateral edges, so thar a straight sharp edge of an angle of 65 deg rees is formed on borh sides. Both sides were weil used, considering the usemarks.

The handaxe Nr. 1, roo, is a fine specimen of almosr planoconvex secrion. The lower face is well-trimmed all over its surface by !arge, shallow primary flaking and by stepflakes along the edges, while the upper part has been reshaped by a few !arge stepflakes which seem to have removed the original, probably pointed apex. The upper face retains a fair amount of cortex in the lower part. The butt has a straight, unifacial edge, while borh lateral edges are shaped by bifacial srepflakes, forming slightly sinuous edges of 70 ro 75 degrees. The rool is slightly weathered and fairly rounded, like Nr. 2 (Fig.8).

The small biface Nr. 3 is sharper and fresher and seems to have been derived from the sediment only recently. Iris a small pointed ovare with lirtle refinement and a thick biconvex secrion. It is made by deep primary flakes and a few stepflakes, so that the edge which runs araund the whole tool is zigzag and sinuous and does not seem ro have been used .

The !arge biface N r. 10 seems tobe unfinished (Fig.9). It retains much correx at the heavy butt and has a thick irregular biconvex section. The tip of the well-made apex is broken off, leaving the rool rather useless. The lateral edges are bifacial but seem ro be unfinished. A heavy percussion mark is seen at the lefr edge, where only a small part of the intended !arge flake broke off, which would have produced a sinuous, usable edge.

Nr. 12, which has come from in situ in the sandstone, is a broken butt-half of a biface and is very fresh indeed. It retains cortex on its upper face next to a few !arge primary flakes . The lower face shows a !arge shallow flake surface with a number of small flakes and stepflakes removed along one edge. Ir has a round butt with a sharp, straight edge, and the left side has a used, partly bifacial , rather straight edge.

Nr. 13 is a fine oval biface, made on a white, fine-grained quartzite cobble or very !arge flake, of which cortex is rerained on both sides and partly on the butt (Fig.10). The lower flar face is trimmed by !arge, shallow flakes and secondary stepflakes ar the right side, raken off from the marginal cortex, so that a unifacial edge is produced. The orher face has only few primary flakes. lt is the poinr which is parcicularly well-shaped, and this must have been the main focus of the tool. Ir is shallow, tongue-like, and very delicare and thin compared to the ~verall heaviness of the tool, with a very straight, sharp edge. The

photo is not able to show the refined apex. Nr. 14 is an elongate ovare, unfinished biface, made from a fine-grained blue-grey quartzite cobble

with a cortical butt. Ir is fashioned only by !arge primary flakes from both sides, forming zigzag lateral edges which have probably not been used. The point has broken away, probably by use, as a !arge, long itudinal flake has removed the original point. Except a few small secondary flakes at the crippled point there is no other secondary retouch . This tool was found in the sand in the recent central g ully bed below the location of the other rools from the hill slope. Other tools from Satpati:

Nr. 9 is either the broken butt parr of a biface or more likely it is a scraper. lt is made on a !arge, split flake where half of the flake broke off during detachment, as the point of percussion is situated at the

The Sa!pati handaxe Jllt and tbe Cbabeni 11111/afe site in sollfhem Ntpal 23

Fig. 7. The butt end of ehe large handaxe Nr. 2

Fig. 8. (lefc and righr): Handaxe Nr. 1 from acpaci Hili .

24 Gudrun Coroin11s

Fig. 10. (lefr and righr): The !arge biface Nr. 13 from Sarpari Hill.

The Satpati handaxt siJe and the Chabeni 11mjace site in Jouthern Nepal 25

Fig. II. A bifacial endchopper from Satpati Hili.

plaee where ehe flake had split imo 2 pieees. Ir has a straight, used working edge opposice ehe poinc of percussion, and is trimmed by small flakes from both faees. 1t is slightly weachered and rounded like ehe bi faees 1 and 2.

Nr. 11 is an incerescing bifaeial endehoppet on an oblong cobble. Je has a bifacial distal edge wich fine, shallow trimming and has an angle of 70 degrees and has been used. Two flakes have been removed from ehe corcex body of ehe cool, a small one on one side and a deep, angular one on ehe other side (Fig.ll).

Nr. 15 is a unifaeial endehopper from an oblong quarczice cobble wich only few unifaeial primary flakes and some secondary recouch and heavy usemarks whieh eaused ehe distal edge to become blunt and coneave. This cool was found cogecher wich ehe 3 flakes and 2 fossil hone fragmencs on ehe hillslope ac ehe ouccrop of ehe lower quarczice gravel lense in ehe sandscone.

Nr. 8 isapick from a splic cobble slicewich jusc a few shallow flakes taken off near ehe poinced apex. This cool does not Iook as if it belongs co ehe same period as ehe ocher palaeolichs, bue seems tobe much younger. It was not found cogecher wich ehe bifaces buc on ehe surfaee in ehe lowesc parc of ehe khola bed.

Nr. 6 is probably ehe broken-off cip of a biface. Ic has bifaeial pr.imary crimming. Ic is a very fresh piece wich a chick calcrece erusc ac one place. Nr. 7 is a bifaeial cool wich a used, bifaeial rounded point, made by unifacial scepflakes. The edge of ehe apex probably eoncinued down along ehe left side, buc a very !arge flake from ehe lower faee has oblireraced most of the lateral edge and seems co have rendered ehe original small biface rather useless.

26 Gudrun Corvinus

Flakes from Satpati: Nr. 5 is a very !arge cortex flake, detached from a very !arge cobble, showing no other trimming.

Nr. 4 is a simple, wide-angled, pointed flake with no Cortex and no trimming. Nr. 16 and 17 aresimple flakes found tagether with flake Nr. 18 and with the endchopper Nr. 15 and

the bone fragments at the outcrop of the lower gravel. The two flakes have been detached from the same

light-grey quartzite core, both showing plain platforms and some lateral cortex and no secondary trimming, but both have a lateral edge, which may have been used.

Nr. 18 is a simple waste flake of quartzite with plain platform and some stepretouch at the platform

edge prior to detachment of the flake. c. Summary of the Satpati cultural remains

Though the cultural material of the Satpati site is not very abundant, it is of great importance due to

the reasons stated in the beginning. The bifaces are made in the Indian tradition and there is no doubt that the connection of the people responsible for the Satpati bifaces is to India. They have crossed the wide Gangeeie Plain and occupied the terrain at the foot of the Himalayan mountains probably during one of

the glacial periods, not rhe last one but probably before the Eemian time. The climare then musr have been cooler and drier and the vegetation less thick than in interglacial and recent times during which the

Himalayan foot was covered by the thick Babai forest zone. After the Satpati handaxe population lived here, the rivers from the north have deposited a few

hundred meter of thick alluvial sediments on top of the site,thereby burying it deeply under their sediments. It is only due to the tecronic uplift and folding that the site has been exposed and has come to

rhe surface again, affering a chance for the searching geo- archaeologist. We must presume therefore, rhat Acheulian man, whose sites we find so abundantly in India south of rhe Gangetic Plains, also occupied

the Gangeeie Plains and that their sires are buried under thick alluvial sediments. Ir also seems obvious, that they ventured further north ro the foot of the Himalayas, as evidenced by the Satpati site, and that they even penetrated right into the Himalayan foothills, as evidenced by the Gadari handaxe site in the Dang Dun valley wirhin the Siwaliks.

These are interesting findings, especially since they are primary, stratified in situ sites, and throw new

light on the extention and migration of Acheulian man. It is not the Gangeeie Plain which formed the northern and north-eastern border of rhe handaxe popularion on the Indian subcontinent, and it is not rhe Himalayan foothills, bur it is the higher Himalayan mountain belt which forms the N and NE boundary

of their migration.

2. The Chabeni culrural assemblage

a. Location

West and south-west of the Satpati Hili there stretches a wide, higher river terrace surface along the

foot of rhe Siwalik hills, which is made up of yellow silt deposits overlain, at places, by colluvial blocks

and boulders of Siwalik sandstone, washed down from the Siwaliks. Its elevation is 15 - 20 m above the lowland plain (see Fig.5). The silt surface of the terrace slopes down to rhe south to a small canal and here breaks up into badlands. Here erosion has set in heavily due to the cutring away of the forest .

At the erosional margin one sees artefacts of quartzite and chert eroding out from the uppermost horizon of the silt at the contact to the overlying colluvial deposit. It must have once been an occupation

site of a !arger camp site, called here the Chabeni site. The artefacts on the surface are spread over an area of about 100 by 100 m . They consist only of stone implements and stone debris and there is no pottery associated with them.

The atpati handaxt Iite and the Chahmi 1111ijace Iile in IOIIthml Nepal 27

Fig. 12. Chabeni site, place 1, covered with arrefacrs, eroding out from rhe silr.

Fig. 13. Chabeni site, place 2, on rhe rerrace surface, wich the iwalik hills in ehe background .

28 Gudrun Corvinm

The terrace was certainly of a much wider extent during the time of the occupation of the Chabeni people, stretching further to the west along the foot of the Siwalik hills, as weil as to the easr along the foot of Satpati Hili. Bur erosion has removed much of its former extension, leaving now only a very reduced terrace surface. The terrace is covered only by brush and an occasional tree. If one proceeds for some few hundred meters on the terrace to rhe east one encounters a !arge area wirh enormaus amounts of potsherds on the surface, but no stone artefacts. Before occupation of rhe present people of Tharus, Kumals (pottery-making community) and Kuwars (idol-making community) who live in rhe recent villages, there musr have existed a potters village north of Chabeni on this rerrace. A few flar ancient-type bricks also can be found here, which are not of the modern type. Ar one place, at a small shrine amongst rhe brush, there are hundreds of broken potsherds especially of fragments of pottery horses. These seem to be still produced here by the Ku war communiry, ordered for special fesrivals to be affered at rhe shrines by the Tharus and Kumals. But these are young archaeological sites and we will deal here only with the Chabeni stone age site.

The sire, as exposed now, is about 100 by 100 m in exrension. Ir has been divided into 4'places' : Place 1,2,3 and 4 are areas wirhin the Chabeni site, each of them not more than 50 by 50 m. Chabeni 1 is the richest place, also called the'flake spot', as here many flakes and waste chips come out from rhe silt (Fig.12). Place 2 is a small area just south of rhe 'flake spot' , where rhe terrace deposits break up into the badlands. Ir is also called the adze place, as here particularly many unifacial adzes have been recorded in 1991 (Fig.13). Place 4 is just northwest ofPlace 2 and west ofPlace 1, and herein 1994 many unifaces have been found. Place 3 is a small area north of Place 4 wirh only a small sample collection. b. The cultural material of the Chabeni site

The collecrions of all the places are sample collections and are rherefore selecrive. An analysis on waste ro tools is not possible and therefore it is uncertain whether the site was a factory and/or camp site. Bur from rhe interpretation of the 53 tools we can deduct nonerheless, that rhe site was a camp- and working place. The artefacts consist in the majority of flakes and stone waste of quartzite and white cherty quartzite and rarely of chert.

Besides the waste there are a number of tools: corescrapers, scrapers and choppers, and particularly frequent are unifaces,which are very special for rhis sire. In fact rhese unifacial tools do resemble Hoabinhian unifacial tools from northern Sourh-East Asia. The collection contains 118 artefacts. They consist of: 38 artefacts from Chabeni, Place 1; 40 artefacts from Chabeni, Place 2; 12 artefacts from Chabeni , Place 3; 27 artefacrs from Chabeni , Place 4;

1 artefact from Chabeni , east ofcanal' . 118 sum Of the 118 artefacts at Chabeni there are:

26 unifaces or unifacial adzes, (all made from cobble); 1 unifacial disc, (from cobble); 1 knife, (from a flake); 7 scrapers, (3 are unifacial, from cobble, 4 are from flake); 5 choppers, (4 are unifacial, from cobble, 1 is partly bifacial); 4 corescrapers, (all 4 are unifacial, from cobble); 1 sling-ball, (from a round cobble); 1 polished,broken cylindrical stone; 7 used flakes;

62 flakes, i .e.waste flakes;

2 flake-like pieces; core;

The Satpati hamlaxe site and the Chubmi 1111i/are site in southem N epal

Fig. 14. (left and right): Uniface Nr. 1 from Chabeni.

29

38 of the 53 tools of the collection at Chabeni are unifacially trimmed cools, that means that 72 % of the cools are unifacial tools made from cobbles or from split cobbles, where one face was left complerely uncrimmed or almost uncrimmed, showing cortex. Tlhis high percemage of unifaces is surprising and is, so far, not found at other sites in Nepal, not even in Patu.

The unifuces were made from quartzite or chert cobbles, from which !arge, flat primary flakes were removed, or they were made from split cobbles. One face was left entirely cortical, or has only a few shallow scars, while the otlher facewas crimmed by shallow flaking along ehe circumference (.Fig.l4). The diagram of Fig.l5 shows that they are of a standardized oval shape, their B/L ratio with a peak ar 0.5 co 0.6 showing that they are neirher very long, nor very short. Their apex is not made inco a straight or oblique straight transversal edge like most Patu adzes, but are unifacially trimmed into a convex apex. They have an amazing resemblance wich Hoabinhian unifacial axes (Fig.16). The length of rhe unifaces is rather unsrandardized (Fig.J 7). Some of the unifaces are very small, only 60-70 mm in length. Orher again are very !arge, up co 175 mm. The thickness ratio of the unifaces is also quite standardized wich medium Th/B values (Fig. l8), wich neither very rhin nor very thick section. Only the few choppers and corescrapers are heavy tools wich rhick section, as is usual for ehern.

30 Gudrtm Co"'irms

30

25

20

15

10

5

0

0,3 0,4 o,s 0,6 0,7 0, 8 0,9 1 1,1

breadth/length ratios

Fig. 15. Diagram of the breadth/length ratios of 27 unifaces from Chabeni .

fig. 16. (left and right): Uniface r. 4 from Chabeni.

% 32

30

28

26

24

22

20

18

16

14

12

10

8

6

4

The Satpati handaxe site and the Chabeni uniface site in southern Nepal

2

0 ,_--~------.---,---,--,---,---.---,---T---+---,---r--,_--.

50 60 70 80 90 100 110 120 130 140 150 160 170 180

length in mm

Fig. 17. Diagram of the length of 27 Chabeni unifaces.

0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8

thickness/breadth ratios

Fig. 18. The thickness/breadth ratios of 27 Chabeni unifaces.

31

32

%

::l 24l 22 20

18

16

14

12

10

8

6

4

Gudrun Corvinus

2

0,_---+----,---.----,---,----,----,---,----,---,---~---,

10 20 30 40 50 60 70 80 90 100 110

length in mm

Fig. 19. Diagram of the length of 68 Chabeni flakes.

0,3 0,4 0,5 0,6 0,7 0,8 o, 9 1 1,1

breadth/length ratios

Fig. 20. Diagram of the breadth/length ratios of 68 Chabeni flakes.

The Satpati handaxe site and the Chabeni uniface site in southern Nepal 33

%

45 ~ 40

35

3oj1

25

I 20 ~

" ~ 10 I

sJ 1

0 0,1 0 ,2 0 , 3 0,4 0,5 0,6 0 ,7

thickness/ breadth ratios

Fig. 21. Thickness/breadrh ratios of 68 Chabeni flakes.

% 90

85

80

70

60

50

40

30

20

10

0

c.p. pl.p und br st no st

left) plotform right) stepretouch at plotform

Fig. 22 The characrerisrics of rhe placform of 68 Chabeni flakes.

34 Gudrun Corvinus

% 90

80

70

60

50

40

30

20

10

3 1,5 0

full 1i opp or no

cortex on dorsal face

Fig. 23. The dorsal cortex on 68 Chabeni flakes.

The few scrapers partly resemble unifaces, as 3 of rhem are made from cobbles and are unifacially trimmed. 4 other scrapers are from flake and are either unifacial, or bifacial.

Of rhe 69 flakes 7 are used flakes. The flakes show no special fearures. Their length ranges from 20 to 100 mm (Fig.19). Their shapes are peaked ar a B/l ratio of0.7, i. e. rhey show lirrle inclination rowards elangare flakes or blades (Fig.20), and rheir thickness/breadth rarios wirb a definite peak ar 0.2 to 0 .4 are rarher reg ular, showing neirher very thin nor very thick flakes (Fig.21).

The flakes of Chabeni show the same characteristics as rhose from Patu (Corvinus 1987). They are struck with 75 % from cortical platforms from the core and not from prepared cores: 75 % have cortex platform, none have a prepared platform and only 9 % have plain platform (Fig.22). And like at Patu, the !arge majority of flakes have steprerouch or stepflaking ar rhe plarform edge on the dorsal face and no orher rrimming (Fig.22). This stepflaking was clone on the core and indicates that the flakes are probably manufacturing flakes of !arger cobble-rools like choppers, coresrapers and unifaces. Surprisingly few flakes are cortex flakes: only 3 % , while 78 % have no cortex at all (Fig.23), apart from rhe cortex p latform.

10.5 % have also some cortex opposite the platform and 1.5 % have cortex along one side like orange slices. There is a great resemblance with the Patu flakes. c. Interpretation of the Chabeni arrefacts

The Chabeni artefact assemblage is a cobble-rool indusrry. The rools have been manufactured from quartzite cobbles by p redominanrly unifacial trimming, leaving one face as a cortical face, so that the edge is a partly cortical edge made by shallow primary and smaller stepflaking to shape the edge from one side only.

The predominant rool type is the uniface, oval or elongate oval in shape, with an edge along rhe whole circumference. A few scrapers, choppers and corescrapers make up the rest of the rools. Flakes are manufacturing debris of tool-making and only few flakes seem ro have been used. The unifacial rools have

The Satpati handaxe site and the Chabeni uniface site in southern N epal 35

a striking resemblance ro Hoabinhian tools from South East Asia. In peninsular lndia such tool types are not encountered and it is possible that the connection of these people was not with lndia, but with South East Asia via North East India. The Chabeni assemblage is rhe first of its kind in Nepal and similar artefacts have so far not been encountered in ocher parrs of N epal.

IV . Concl usions

The area of Satpati/Chabeni in the Lumbini zone in souehern Nepal has yielded two cultural assemblages, one from the palaeolithic and one from the mesolithic period.

The palaeolithic site of Satpati is situated in a small hill, the Satpati Hill, at the foot of the Himalaya, bordering the Gangeeie Plain. Early palaeolithic bifacial rools of bifaces have been discovered wirhin alluvial sediments in this hill, belanging to the alluvial body of sediments of the Gangeeie Plain, which, at this particular place, have been subjected ro tectonic uplift and folding by the activiry of the Himalayan Frontal Thrust zone. The tools had been embedded in the Gangeeie alluvium in the later Middle Pleisrocene, probably more than a 100 000 years ago, and had been buried by subsequent alluvial deposits of probably more than 200 m thickness . Normally they would have remained buried for ever, if not recent tectonic activities at the Himalayan foot had uplifted these deposits in this particular small place, just where by chance the handaxe site was embedded, and had exposed these deposits by thrusting them over the youngest alluvial sediments of the recent Gangeeie Plain.

This is a unique find in a unique situation. Normally much older rocks of the Tertiary Siwalik Group of sediments of the Siwalik Hills are thrust over the Gangeeie alluvium at the Himalayan mountain front . Bur at this particular place at Satpati, rhe frontal thrust bifurcates into two strands: the older northern Himalayan Frontal Thrust thrusting Siwalik rocks over the Gangeeie alluvium, and the younger souehern thrust uplifting and folding a small slice of the older Gangeeie alluvium and thrusting this slice over younger Gangeeie alluvium. And by chance an occupation place of handaxe people was embedded in this small slice of older alluvium and thus got exposed by this recent tecronic phenomenon. The handaxe population lived here at the Himalayan foot in the plains, after crossing the wide Gangeeie Plain, but they certainly belonged ro the people of Acheulian rradition from lndia.

Much later, probably in early Holocene times, much younger people came to live here almost at the same place at the mountain front, settling on the higher alluvial terrace remnants north of the neighbouring Chabeni village. They produced unifacial rools made from cobbles, which seem to have no traditional affinities to the prehistorical culrures of peninsular India soueh of the Gangeeie Plain. They instead seem ro have closer affinities ro cultures which lie to the south east and east, i . e. ro industries akin to the Hoabinhian indusrry of Soueh East Asia.

Further to the east in Nepal the author had discovered in 1985 another very particular cultural

assemblage, the Patu industry at Patu at Raro Khola, (Corvinus 1987), with unifacial and bifacial adzes which also seems ro have affinities ro South East Asia rather than ro India. Further to the west in Nepal, where the author is working extensively in the Dang-Deokhuri Dun valleys wirhin the Himalayan foothills, a great nurober of prehisroric sites could be recorded, from the early palaeolithic period (wich a first handaxe assemblage discovered in Nepal from basal deposits of the Dun alluvium, Corvinus 1991, 1993) via late Pleistocene flake and chopper industries and microlithic industries to neolithic remains (Corvinus 1985, 1994) which all have doubtless connections with the prehistoric traditions of rhe Indian peninsula.

Research in Nepal is continuing by the author in order ro further unravel the intricate and intriguing patterns of the prehistoric occupations along the Himalayan mountain front .